Data Coverage Claim Verifier
Pricing
$32.50 / private audit
Tests a vendor's explicit coverage claim against an authorized buyer reference set and returns PASS, FAIL, or UNKNOWN with reproducible evidence.
free_example ignores all private audit fields and uses bundled synthetic data. private_audit validates and audits your manifest.
The quoted claim and the numerical threshold being tested. Only at_least coverage claims are supported.
Declare whether the supplied reference is the complete buyer-defined set or a representative sample of a larger population.
An inline record array, an explicitly allowlisted HTTPS JSON/JSONL/CSV URL, an Apify Dataset, or an Apify key-value record.
An inline record array, an explicitly allowlisted HTTPS JSON/JSONL/CSV URL, an Apify Dataset, or an Apify key-value record.
Parallel referenceKeys and vendorKeys arrays define one exact or composite join key. No fuzzy or AI matching is performed.
Up to five reference fields to report separately, for example market or company size.
Cap tested reference records and set the stable sampling seed. Maximum 20,000 tested records.
Wilson confidence intervals apply to representative or downsampled tests; complete reference sets are measured directly.
You must have the right to audit both supplied sources, must not bypass access controls, and must minimize the data supplied.
Identifiers are hashed by default. Raw join keys require explicit counterexample-output authorization.
Optional loss thresholds used only for the separate buy/test_more/negotiate_or_discount/reject suggestion.
A non-personal datasetVersion label supports version-repeat measurement; only its SHA-256 fingerprint is output.
